A revolutionary security system is undergoing a two-year testing phase at the Blue Grass Army Depot in Kentucky.

The U.S. Army’s Physical Security Enterprise and Analysis Group provided the funding for the project initiated by the Directorate of Emergency Services at BGAD, and the U.S. Army Corps of Engineers’ Mandatory Center of Expertise is collaborating with BGAD for it.

The security system is a commercial-off-the-shelf proprietary product that is a real-time modular physical threat detection solution using artificial intelligence and deep neural learning computer vision to modernize existing electronic security systems by detecting objects/intruders, guns, fights, slips and falls, smoke and fire, facial recognition, and behavior anomalies of individuals.

The product processes camera video stream content and sends alerts to security response forces according to configurable and customizable protocols, and officials at BGAD, a subordinate of the Joint Munitions Command, are impressed with its capabilities.

“It’s a great tool that is modernizing existing technology into our force protection posture. This is going to be a game changer for security and protection,” said Jim Vaughn, JMC’s provost marshal. “The Army is looking to use current technology to replace and or backstop antiquated security systems.”

Toward the end of January, BGAD and the makers of the security system software held a demonstration of its capabilities, and the first involved an active shooter scenario in real time using the system for gun (pistol and long rifle designation) detection, facial recognition, vehicle tracking, and fight detection using BGAD’s existing closed-circuit television cameras.

“It provides an artificial intelligence/machine learning capability that the existing cameras don’t have,” Vaughn said. “It’s basically a software package that integrates with existing CCTV cameras.

“Weapon detection typically occurred within seconds in nearly every case,” he added. In a few instances, the longest detection time for weapons was six to eight seconds.”

The second demonstration involved using the artificial intelligence via CCTV for intrusion detection and perimeter protection. Detections occurred when individuals or vehicles entered a polygon shaped detection zone that was configured by BGAD using the system’s graphical user interface dispatch headend.

“The system is able to detect from hundreds of meters away when someone walks into the perimeter,” Vaughn said. “It can detect if it’s an animal, say a deer, or a person, and it shows where the person is at, and gives a description of what they’re doing. The system can also be trained to look for mechanical flying devices and natural (birds) flying devices.

“Currently, it’s achieving around 97% accuracy in testing, while also ensuring that the system meets requirements for minimizing nuisance and false alarms. This entails the system detecting only standard intruders, various types of guns, smoke, fire, and so forth,” Vaughn added.

These efforts support JMC’s Campaign Plan initiatives to enhance security and protection of the Joint Strategic Support Area by closing potential security gaps at JMC’s installations.